INDEX

Explanations

the name "Julie" or related forms

oai_token-act-pair · gpt-3.5-turbo

mentions of specific names

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

asking

-0.77

otin

-0.76

izoph

-0.75

alez

-0.75

itutional

-0.74

iment

-0.72

 bearer

-0.71

 retard

-0.69

opic

-0.68

mental

-0.68

POSITIVE LOGITS

 Julie

1.11

 Andrews

0.90

 Walters

0.89

 Bowen

0.85

 Klaus

0.85

 Johnston

0.84

 Ruin

0.83

 Wallace

0.80

 Bind

0.80

 McMahon

0.78

Activations Density 0.008%

the name "Julie" or related forms

mentions of specific names

No Comments

No Known Activations